Substance use-related harm among college students remains an important national problem associated with both morbidity and mortality. Previous research has identified validated screening instruments for substance use and multiple risk factors for engaging in substance use. Despite significant resources devoted to these efforts, alcohol use and harm rates on college campuses have not decreased. At present, a gap exists in translating factors from behavioral models into effective interventions to reduce substance use among college students. Our long-term goal is to use social networking web sites (SNSs) to identify substance use in college students and provide feasible, scalable and targeted interventions. Before such interventions can take place, however, further evaluation regarding SNSs as an instrument for identifying substance use and as a mediator within mechanisms of behavior change is warranted. Approximately 97 percent of undergraduates report ownership of a Facebook (FB) profile and display of references to substance use is common. These references are therefore both created and consumed by college students, and may play a key role in mediating behavior. The objective of this application is to test the predictive value of FB as a substance use identification tool, evaluate a provisional model of FB as a mediator of behavior change, and explore methods by which FB can be used as a targeted intervention tool. Our central hypothesis is that displayed substance use behavior on FB will be a highly sensitive tool for the identification of substance use as well as a mediator for behavior change related to substance use. The proposed study will identify an inception cohort of 600 college freshmen from the University of Wisconsin-Madison and the University of Washington. Incoming freshmen will be recruited via email and phone; consented participants will be friended on Facebook (FB). Data collection will include monthly FB profile evaluation for substance use. We will conduct biyearly interviews using new technology measuring substance use behaviors, and yearly interviews to evaluate our provisional model of FB as a mediator in behavior mechanisms of social transmission. Among participants who display substance use references indicating behavior change, at the time the reference is identified on FB we will conduct an additional interview using new technology to investigate the role of FB as a mediator in the behavior change. We will also conduct focus groups to explore the feasibility of and barriers to FB as an intervention tool. Using FB as our laboratory and participant self-report as our field work we will determine the role of FB as an instrument for identifying substance use and test a provisional model of FB as a mediator in behavior change. Given the popularity and ubiquity of FB among the college population, this highly innovative tool has potential to improve universities' capacity to provide interventions that are feasible, scalable and targeted.